MONONUCLEAR PHAGOCYTES (KUPFFER CELLS)
AND ENDOTHELIAL CELLS
Identification of Two Functional Cell
Types in Rat Liver Sinusoids
by Endogenous Peroxidase Activity
JEAN-JACQUES WIDMANN, RAMZI S . COTRAN, and
H . DARIUSH FAHIMI
From the Harvard Pathology Unit, Mallory Institute of Pathology at Boston City Hospital,
Boston, Massachusetts 02118
ABSTRACT
The fine structural characteristics and phagocytic properties of peroxidase-positive and
peroxidase-negative cells in rat hepatic sinusoids were investigated . Cells with a positive
peroxidase reaction in the endoplasmic reticulum and the nuclear envelope make up approximately 407, of cells in rat hepatic sinusoids and have abundant cytoplasm containing
numerous granules and vacuoles, and occasional tubular, vermiform invaginations . After
intravenous injection of colloidal carbon, the luminal plasma membrane of these cells
shows continuous sticking of carbon, and there is evidence of avid phagocytosis of colloidal
carbon particles . Peroxidase-positive cells are the only cells in hepatic sinusoids which
phagocytize large (0 .8 µ in diameter) latex particles . In contrast, the peroxidase-negative
endothelial cells, which make up 48% of cells, have scanty perinuclear cytoplasm and
organelles, and their long cytoplasmic extensions that form the lining of the hepatic sinusoids have fenestrations ; these cells ingest small amounts of colloidal carbon, principally
by micropinocytosis, exhibit no sticking of carbon particles to their plasma membranes,
and do not ingest the larger (latex) particles . The so-called fat-storing cells are peroxidase
negative and totally nonphagocytic . The peroxidase reaction thus distinguishes the typical
mononuclear phagocytes or Kupffer cells of rat liver from the endothelial-lining cells .
Since the initial observation of phagocytic activity
in cells lining the hepatic sinusoids (1), and the
description of phagocytic Kupffer cells by Von
Kupffer and others (2-4), the question : "Whether
hepatic sinusoids are lined by a single cell type or
by two or more different distinct types of cells"
has been the subject of controversy . Several comprehensive reviews of the literature have noted
that since all hepatic sinusoidal cells are capable
of a certain degree of phagocytosis, they must be
THE JOURNAL OF CELL BIOLOGY • VOLUME 52, 1972 .
functional variants of the same single cell type
(5-7) . Although several recent fine structural
studies (8-12) indicate the presence of at least
two types of lining cells, i .e . a wall-forming
endothelial cell and a phagocytic Kupffer cell,
other authors emphasize that there are numerous
intermediate forms and, on ultrastructural grounds
alone, raise again the question of two variants of
the same cell type (6) .
In a recent cytochemical study from this labora-
pages
1 5 9 -170
159
tory, Fahimi reported that approximately 40 %
of sinusoidal-lining cells in rat liver showed a
characteristic pattern of endogenous peroxidase
activity in the endoplasmic reticulum and the
perinuclear cisternae (13) . A similar pattern of
localization of peroxidase has also been described
in peritoneal macrophages of guinea pigs and rats
by Cotran and coworkers (14, 15) . The present
study was undertaken in order to determine
whether such peroxidase activity in hepatic sinusoidal-lining cells could be correlated with distinct
morphological and functional (phagocytic) properties and thus could be used as a marker to
differentiate Kupffer cells from endothelial cells .
The results indicate that peroxidase-positive
cells, as compared to the peroxidase-negative
cells, (a) ingest significantly larger amounts of
relatively small particles (carbon 0 .02-0.05 µ),
(b) exclusively phagocytize larger particles (latex
0 .8 is), and (c) possess morphological characteristics that have been ascribed by others (8-12)
to typical Kupffer cells. The peroxidase reaction
can thus clearly differentiate two types of sinusoidal-lining cells : Kupffer cells and endothelial
cells .
MATERIALS AND METHODS
Cytochemical Studies
Normal female adult albino rats (Charles River
strain) weighing 200-250 g each and fasted for 24 hr
were used . The liver was fixed by perfusion through
the portal vein (16) with 2 or 2 .25% purified glutaraldehyde (17), in 0 .1 M sodium cacodylate buffer,
pH 7 .3, for approximately 10 min . After several
rinses in the same buffer, 40- s-thick sections were
cut on a Sorvall TC2 tissue sectioner (Ivan Sorvall
Inc ., Norwalk, Conn .) (18), collected in 0 .05 M TrisHCl buffer, pH 7 .6, and incubated in a modified
Graham and Karnovsky (19) medium containing 0 .1% 3 , 3'-diaminobenzidine-tetrahydrochloride
(DAB) (Sigma Chemical Co., St . Louis, Mo .), and
0 .01 % H202 in 0 .1 M Tris-HCl buffer, pH 7 .6, for
60 min at room temperature . After 90 min of postfixation with 2% osmium tetroxide in distilled water
(20), sections were dehydrated in graded ethanol
solutions and embedded in Epon 812 (21) . 1-is-thick
sections were examined by light microscopy both
unstained and stained with 1 % toluidine blue .
Thin sections for electron microscopy were cut on an
LKB ultrotome III (LKB Instruments, Rockville,
Md .) and examined either unstained or lightly counterstained with lead citrate in a Philips EM 200 electron microscope . Cytochemical controls for the
peroxidase activity were performed by incubation of
160
sections from all experimental animals in media in
the absence of hydrogen peroxide.
Phagocytosis Experiments
COLLOIDAL CARBON : Under light ether anesthesia, animals were injected intravenously with 0 .05
ml/100 g weight of filtered, shellac-free, nontoxic
colloidal carbon (Gunther Wagner, Hannover,
Germany, Lot c-11-1431/a, containing 100 mg
carbon/ml), and the livers were fixed by perfusion at
2, 5, and 30 min . Such carbon particles measure approximately 0 .03 µ4 in diameter (22) .
LATEX : In another group of animals, 0 .05
ml/100 g weight of latex particles with mean diameter of 0 .8 µ (Bacto-Latex 0 .81, code 3102 from
Difco Laboratories, Detroit Mich .) were injected
directly into the portal vein of anesthetized and
laparotomized rats ; the livers were fixed by perfusion
2 min and 5 hr after the injection of particles .
RESULTS
Fine Structural Cytochemical Characteristics
of Various Sinusoidal-Lining Cells
On the basis of ultrastructural characteristics
and peroxidase activity, three different types of
cells could be differentiated (Fig . 1)
(a) The typical peroxidase-positive cells (KC
in Fig . 1) exhibited electron-opaque reaction
product in the nuclear envelope and segments of
the endoplasmic reticulum . These cells had relatively abundant cytoplasm, and were rich in
cytoplasmic organelles, and had numerous granules of varying density and shape . Peroxidasepositive cells were seen overlying the extensions
of typical peroxidase-negative endothelial cells
and were found occasionally in direct contact
with microvilli of hepatocytes .
(b) The typical wall-forming endothelial cells
(EC in Fig . 1) were peroxidase negative, possessed
a smaller nucleus, and had less cytoplasm and
fewer organelles. The cytoplasmic extensions of
these cells showed many micropinocytotic vesicles
and, in proper sections, exhibited fenestrations or
sieve plates (10) (Figs . 1, 5, and 7) . In some areas,
there were larger discontinuities in the endothelium as described by several investigators
(23, 47), but we could not determine whether
such discontinuities were preexisting gaps which
increased in size owing to high perfusion pressure .
This matter was not pursued further in the present
study.
(c) The third cell type was the so-called fat-
THE JOURNAL OF CELL BIOLOGY . VOLUME 52, 1972
FIGURES 1-7 are from livers of rats injected with colloidal carbon ; the livers were fixed at different time
intervals by perfusion and processed for peroxidase activity . All sections were counterstained with lead
citrate. Length of scale line for all figures is 1 µ .
This electron micrograph illustrates the three types of cells encountered in rat hepatic sinusoids . A Kupffer cell (KC), which is seen at the bifurcation of the sinusoid (SIN), shows peroxidase
reaction product in the perinuclear cisternae and ER . A typical wall-forming endothelial cell (EC),
elongated in shape, shows no evidence of staining for peroxidase . A "fat-storing cell" (FS C) contains a
few fat droplets and lacks peroxidase reaction product . Note that fine extensions of an endothelial cell
cover the luminal surface of the fat-storing cell (arrows), indicating the true perisinusoidal location of
such cells . Kupffer cells and endothelial cells are thus the only two cell types which line the sinusoids .
Hepatocytes, H ; collagen, CO . X 5700 .
FIGURE 1
WIDMANN ET AL .
Endogenous Peroxidase: Marker for Kupffer Cells
161
storing cell (FCS in Fig . 1) (24-26) . These cells
were characterized by the presence of a few lipid
droplets, a few mitochondria, and a prominent
granular endoplasmic reticulum . They were
always peroxidase negative (Fig . 1) . Such cells
were separated from the sinusoidal lumen by
thin extensions of endothelial cell cytoplasm (Fig .
1), and thus can be considered as truly perisinusoidal cells .
Of over 1000 sinusoidal-lining cells counted,
39 17c of cells were peroxidase-positive Kupffer
cells, 48% endothelial cells, and 13 0/0 fat-storing
cells .
Uptake of Colloidal Carbon by
Sinusoidal-Lining Cells
Light microscope examination of thick Epon
sections from livers of animals injected with 0 .05
ml/100 g weight of colloidal carbon revealed that
larger amounts of carbon particles were picked
up by sinusoidal-lining cells around the portal
tract as compared to those in the center of hepatic
lobules . Studies with electron microscopy revealed
a striking difference in distribution of injected
carbon particles between peroxidase-positive and
peroxidase-negative cells (Fig . 2) .
PEROXIDASE-POSITIVE CELLS : 2 and
5
min after the injection of carbon particles, the
peroxidase-positive cells in the periportal region of
hepatic lobules exhibited a continuous line of
carbon particles attached to the cell membrane
facing the luminal aspect of the hepatic sinusoid
(Fig . 2) . A narrow space of approximately 150200 A separated the carbon particles from the cell
membrane of the peroxidase-positive cells (Figs .
2 and 3) . Occasional pseudopods projecting from
the luminal surface membrane of such cells also
showed evidence of sticking of carbon . At these
intervals, fairly large numbers of carbon particles
were phagocytized by peroxidase-positive cells
and were contained in membrane-bounded
phagosomes and phagolysosomes close to the
luminal surface of the cell . 30 min after injection,
there was no more evidence of carbon sticking on
the surface of peroxidase-positive cells, but the
particles appeared in large vacuoles located deep
in the cytoplasm (Fig . 4) close to the nucleus .
Occasional tubular structures beneath the luminal
surface of peroxidase-positive cells, the so-called
"micropinocytosis vermiformis" (27, 28), also
contained carbon particles at 30 min (Fig . 4) .
162
PEROXIDASE-NEGATIVE CELLS : Of the two
other cell types which border the hepatic sinusoids
and which are peroxidase negative, the fat-storing
cells never showed any evidence of uptake of
carbon particles . The endothelial cells, however,
did ingest carbon particles but considerably less
so than the peroxidase-positive cells (see quantitative data below) . 2 and 5 min after the injection,
there was no linear sticking of carbon to the luminal surface membrane . Fig . 2 demonstrates the
contrast between a peroxidase-positive cell with
carbon sticking to the luminal surface and a
peroxidase-negative cell without such linear
sticking of carbon . Occasional micropinocytotic
vesicles containing carbon particles were seen
below the surface membrane of peroxidasenegative cells (Figs . 2, 5, and 6) . Rarely, larger
intracytoplasmic vacuoles with an approximate
diameter of 0 .3-0 .5 s were also seen to contain
carbon (Figs . 2 and 6) . In comparison to vacuoles
of about the same size in peroxidase-positive cells,
the vacuoles of peroxidase-negative cells contained consistently fewer particles of carbon,
which were distributed at the periphery of such
vacuoles such that the center of the vacuoles was
electron lucent (Fig . 6) . No marked difference
was noted in the amount of carbon in vacuoles of
peroxidase-negative cells at 5 and 30 min after
the injection .
QUANTITATIVE DATA ON CARBON PHAGO -
In an attempt to obtain a quantitative estimate of the difference in phagocytic
activity between peroxidase-positive and peroxidase-negative sinusoidal-lining cells, we examined
approximately 1000 cells in sections prepared
from different blocks of animals injected with
carbon and sacrificed at 2 and 5 min . Peroxidasepositive and -negative cells were divided arbitrarily into four categories (0 to 3+) according to
the number and size of carbon-containing vacuoles
in each cell type . The results of such counts are
presented in Table I . It can be seen that greater
than 1 + carbon was seen in 40 % of peroxidasepositive cells in contrast to only 5 % of peroxidasenegative cells . The relatively large proportion of
peroxidase-positive cells exhibiting no phagocytosis
is believed to be due to the small dose of carbon
used in these experiments ; such cells were seen
principally in the central part of the lobules .
These quantitative results indicate that although
peroxidase-negative cells are able to ingest small
amounts of carbon particles, most of the cells
CYTOSIS :
THE JOURNAL OF CELL BIOLOGY • VOLUME .5e, 1972
FIGURE 2 Electron micrograph of hepatic sinusoid 5 min after injection of carbon, showing the linear
sticking of particles on the luminal surface of a peroxidase-positive Kupffer cell (KC) and the contrasting absence of such sticking on the endothelial cell (EC) . In the Kupffer cell, carbon particles are also contained in phagosomes situated close to the luminal cell surface, whereas the endothelial cell shows only
scant particles in micropinocytotic vesicles (MPV) or in larger intracytoplasmic vacuoles (V) . The
Kupffer cell is separated from the hepatocellular microvilli by discontinuous endothelial cytoplasmic
processes (E), but also shows occasional direct contact with the hepatocyte (arrow) . Sinusoidal lumen,
sinusoid . X 9800 .
WIDMANN ET AL .
Endogenous Peroxidase : Marker for Kupffer Cells
1 63
FIGURE 3 Electron micrograph of the junction (arrow) between a peroxidase-positive Kupffer cell
(KC) and a cytoplasmic process of an endothelial cell (E), 5 min after injection of carbon . Sticking particles are exclusively present on the luminal surface of the Kupffer cell . Carbon-containing phagosomes
are found close to the cell surface . Sinusoidal lumen, SIN. X 15,600 .
FIGURE 4 High-power view of a peroxidase-positive Kupffer cell 30 min after injection of carbon particles . Carbon is found in large phagosomes and phagolysosomes deep in the cytoplasm, and in the "micropinocytosis vermiformis" (arrows) close to the luminal cell surface . X 17,900.
1 64
THE JOURNAL OF CELL BIOLOGY . VOLUME 52, 1972
FIGURE 5 Part of a peroxidase-negative endothelial cell (EC) 2 min after injection of carbon . A few
carbon particles are found on the cell surface (C) or in micropinocytic vesicles (MPV) . X 22,600.
FIGURE 6 High-power electron micrograph of an endothelial cell 5 min after injection of carbon, showing a large carbon-containing cytoplasmic vacuole (V) . The particles are at the periphery of the vacuole.
A few carbon particles are also found in micropinocytic vesicles (MPV) . nucleus, N . X 37,000.
FIGURE 7 High-power view of a "sieve plate" (arrows) (10) in an endothelial-lining cell overlying a
"fat-storing cell" (FSC) . Sinusoidal lumen, SIN. X 24,200.
165
TABLE I
Phagocytosis of Colloidal Carbon by PeroxidasePositive and Peroxidase-Negative Hepatic
Sinusoidal-Lining Cells
Peroxi- Peroxi- Fatdase- Base- storing
positive negative
cells
No . of cells counted
Quo with no phagocytosis
% with + carbon
ofo with ++ carbon
ojo with +++ carbon
411
33
27
25
500
56
39
5
143
100
0
0
15
0
0
with large phagosomes and phagolysosomes containing carbon particles (3+) are the peroxidasepositive cells . Of 143 fat-storing cells examined,
none showed any evidence of ingestion of particles .
Also, it is to be emphasized here that the quantitative difference between peroxidase-positive and
-negative cells as regards linear carbon sticking
(Fig . 2) was absolute . Only peroxidase-positive
cells exhibited this phenomenon .
Phagocytosis of Latex Particles
2 min after the injection of latex particles
(mean diameter 0 .8 µ) into the portal vein, a
large number of such particles were seen phagocytized exclusively by peroxidase-positive cells
in periportal areas of hepatic lobules . The particles
were occasionally noted to be sticking to the cell
surface or, more frequently, they were trapped
between the pseudopods extending from the surface membrane into the lumen of the hepatic
sinusoids (Figs . 8 and 9) . In addition, some
particles were found in phagosomes inside the
peroxidase-positive cells. Of more than 100
peroxidase-negative endothelial-lining cells examined, none showed phagocytosis of latex particles . 5 hr after injection, only peroxidase-positive
cells exhibited intracytoplasmic latex .
DISCUSSION
The studies presented in this paper indicate that
rat hepatic sinusoidal-lining cells can be differentiated, on the basis of peroxidase reactivity,
into two major cell types (excluding "fat-storing
cells") with different functional and morphological
characteristics : the peroxidase-negative endothelial-lining cell and the peroxidase-positive
phagocytic Kupffer cell .
Aschoff in 1924 grouped together several kinds
166
of cells with differing degrees of phagocytosis,
such as endothelial cells, fibrocytes, reticular cells
of spleen and lymph node, hepatic sinusoidal
cells, histiocytes, and monocytes, and designated
this grouping the "reticuloendothelial system"
(RES) (29) . He admitted, however, that vascular endothelial cells and fibrocytes should be
separated because of their low phagocytic activity, and he stated that the concept of the
"reticuloendothelial system" did not imply that
all of the cells were identical . The definition of
the RES as used by Aschoff was based on the uptake of vital dyes which he believed were taken
into the cells by the process of phagocytosis . Lewis
in 1931 described pinocytosis (30) (also called
micropinocytosis [31]) as the process of uptake of
fluid droplets into the cells, in contrast to phagocytosis which is the ingestion of solid particles . Recent studies suggest that these two processes may
have certain common as well as distinct features
(32, 33) .
In mammals, micropinocytosis is observed in a
variety of cell types, including vascular endothelial
cells (34, 35), but avid phagocytosis is limited to a
small group of specialized cells (36) . In a recent
comprehensive review (36), the term reticuloendothelial system (RES) has been replaced by
mononuclear phagocyte system (MPS), a term
which is based on contemporary knowledge of
morphological and functional characteristics of
these cells and their precursors in bone marrow
and peripheral blood (36) . The two most important functional criteria that justify inclusion of
these cells in a single cell system are : (a) avid phagocytosis and (b) firm adherence to a glass surface .
In this context, our studies with particles of different sizes clearly demonstrate that, whereas the
uptake of colloidal carbon occurs in both types of
sinusoidal lining cells, it is much greater in the
peroxidase-positive cells, which, in turn, are also
the only cells capable of phagocytosis of larger
latex particles . That regular peripheral vascular
endothelium can ingest small amounts of colloidal
carbon under some circumstances has been clearly
shown by electron microscope studies after carbon
overloading (22) . The linear sticking of carbon
particles to peroxidase-positive cells may reflect
the sticky nature of the plasma membrane of these
cells . Such stickiness of the surface of some but
not all hepatic sinusoidal-lining cells was first
observed by light microscopy in 1926 by Pfuhl
(37) and was confirmed by electron microscopy in
THE JOURNAL OF CELL BIOLOGY • VOLUME 52, 1979
FIGURES 8 and 9 are from livers of animals injected with latex particles ; the livers were fixed 2 min later
by portal vein perfusion and processed for peroxidase activity . Sections were counterstained with lead
citrate.
FIGURE 8 Electron micrograph of a peroxidase-positive Kupffer cell 2 min after portal injection of latex
particles . Latex particles (L) are found in phagocytic vacuoles or on the cell surface (arrow) . X 11,200 .
FIGURE 9 An avidly phagocytizing Kupffer cell, 2 min after injection . Latex particles are seen in phagocytic vacuoles, some on the cell surface where they stick to the luminal cell membrane or appear partially
surrounded by pseudopod formations (arrow) . X 15,600.
1 67
1958 by Parks (38), and more recently by Wisse
and Daems (12) . The avid phagocytosis and the
sticky plasma membrane of peroxidase-positive
cells thus indicate that the peroxidase reaction is
present exclusively in those hepatic sinusoidallining cells that are members of the mononuclearphagocyte system.
The fine structural characteristics of peroxidasepositive and -negative cells in rat liver, as describ°d in this study, also correspond to the morphological criteria outlined recently by Wisse
(10) and Wisse and Daems (11, 12) and other
investigators (8 ,9, 26) for distinguishing mononuclear phagocytes (Kupffer cells) from the endothelial cells in sinusoids of rat liver. These morphological criteria, such as the amount of cytoplasm,
location of the cell, presence or absence of fenestrations, and number of dense bodies, require that a
large representative portion of a given cell be
present in a section, and this involves a more or
less subjective evaluation of several variable factors . For example, dense bodies may be absent in
part of an otherwise typical Kupffer cell, whereas
some endothelial cells may show a few dense
bodies . In our experience, topography is not an
easy criterion to assess ; Kupffer cells are often
separated from the perisinusoidal space by cytoplasmic processes of wall-forming endothelial
cells, but they may also show some focal direct contact with hepatocellular microvilli (Fig . 2) and,
in the periportal regions of hepatic lobule, they
may be found in the perisinusoidal space itself .
Endogenous peroxidase activity appears to be a
reliable and easy criterion for distinguishing between Kupffer cells and endothelial cells . Our
study shows that a peroxidase-positive cell is characterized by staining of its entire nuclear envelope
and endoplasmic reticulum. No intermediate
forms of staining were ever encountered . Further,
the widely distributed endoplasmic reticulum
throughout the cytoplasm provides labeling even
if a very small part of a Kupffer cell is present in a
section .
The possibility that the endothelial cells are
either precursors or mature variants of the peroxidase-positive Kupffer cells cannot be completely
excluded from this study . It seems highly unlikely,
however, since, as emphasized above, intermediate
forms were not encountered . The origin of Kupffer
cells is not entirely clear : although derivation from
lymphocytes has not been completely excluded,
the balance of evidence suggests a bone-marrow
168
monocyte origin (36) . In this respect, the similarity of the peroxidase reaction of Kupffer cells to
that of peritoneal macrophages (14, 15) is also
more in accord with a monocytic ancestry for
Kupffer cells .
At present, the functional significance and the
exact chemical nature of the peroxidase reaction
in Kupffer cells are unknown . A similar pattern of
enzyme localization has been observed in peritoneal macrophages of guinea pig (14) and rat
(15), in developing granulocytes (48-50), and in
epithelial cells of uterus (39), colon (40), and
salivary (41), lacrimal (42), and thyroid glands
(51) . Studies of Klebanoff and others (43-46) indicate that peroxidase from several of these tissues,
in association with hydrogen peroxide and a
halide, can exert a strong bactericidal, virucidal,
and fungicidal activity . It is conceivable that
Kupffer cell peroxidase may also participate in a
bactericidal mechanism such as the one suggested
by Klebanoff for other peroxidases ; our observation that the peroxidase reaction is limited to
avidly phagocytic Kupffer cells, which are known
to play an important role in the clearing of living
microorganisms from the blood, would be consistent with such a notion .
Whatever its function, however, the peroxidase
reaction in Kupffer cells has clearly documented
the existence of two functionally and morphologically different cell types in the lining of the
hepatic sinusoids and thus could prove useful as a
marker in studies of Kupffer cell kinetics .
The technical assistance of Miss Beatrice Anne Gray
and secretarial aid of Miss Tamara Jane Cubi is
gratefully acknowledged . This study was supported
by grant number 08533 from the National Institute
of Neurological Diseases and Stroke, and HE 08251
from the National Heart Institute, National Institutes of Health, Bethesda, Maryland 20014 .
Dr . Widmann is a recipient of a fellowship from
the World Health Organization . Dr. Cotran and
Dr . Fahimi are recipients of Research Career Development Awards from National Institutes of Health,
Bethesda, Maryland, 20014.
Received for publication 28 June 1971, and in revised
form 30 August 1971 .
REFERENCES
1 . HOFFMANN,
TnE JOURNAL OF CELL BIOLOGY . VOLUME .5h, 1972
F. A ., and F .
VON RECKLINGHAUSEN .
1867 . Ueber die Herkunft der Eiterkörperchen .
Centralbl. Med. Wissensch . 5 :481 .
2 . PoNFICK, E . 1869 . Studien ueber die Schicksale
körniger Farbstoffe im Organismus . Arch .
Pathol . Anat . Physiol. Klin . Med. (Virchows) .
48 :1 .
3 . VON KUPFFER, C . 1876. Ueber Sternzellen der
Leber, Briefliche Mittheilung an Prof. Waldeyer. Arch . Mikrosk . Anat . Entwicklungsmech .
12 :353 .
4. VON KUPFFER, C . 1899 . Ueber die sogenannten
Sternzellen der Säugethierleber . Arch . Mikrosk .
Anat . Entwicklungsmech . 54 :254.
5. ATERMAN, K . 1963 . The structure of the liver
sinusoids and the sinusoidal cells . In The
Liver, Morphology, Biochemistry, Physiology .
C . Rouiller, editor. Academic Press Inc .,
New York . 1 :61 .
6 . NICOLESCU, P., and C. ROUILLER. 1967. Beziehungen zwischen den Endothelzellen der
Lebersinusoide and den von Kupfferschen
Sternzellen . Eleckronenmikroskopische Untersuchung. Z. Zellforsch . Mikrosk . Anat . 76 :313 .
7 . BENACERRAF, B . 1963 . Functions of the Kupffer
cells . In The Liver, Morphology, Biochemistry,
Physiology. C . Rouiller, editor . Academic
Press Inc., New York . 2:37.
8 . YAMAGISHI, M . 1959 . Electronmicroscopic studies
on the fine structure of sinusoidal wall and
fat storing cells of rabbit livers. Arch. Histol .
Jap . 18 :223 .
9 . TANIKAWA, K ., K . YOSHIMURA, and S . GOHARA .
1965 . Fine structure of the reticuloendothelial
cells in the normal rat liver . Morphological
classification . Kurume Med. J. 12 :139 .
10 . WISSE, E . 1970. An electron microscopic study
of the fenestrated endothelial lining of rat
liver sinusoida . J. Ultrastruct. Res. 31 :125 .
11 . WISSE, E ., and W. T. DAEMS . 1970 . Fine structural study on the sinusoidal lining cells of rat
liver . In Mononuclear Phagocytes. R. van
Furth, editor . F . A . Davis Co., Philadelphia,
Pa. 200.
12 . WISSE, E ., and W. T. DAEMS . 1970 . Differences
between endothelial and Kupffer cells in rat
liver . Proc. 7th Int . Congr . Electron Microsc .
3 :57.
13 . FAHIMI, H . D . 1970. The fine structural localization of endogenous and exogenous peroxidase
activity in Kupffer cells of rat liver . J . Cell
Biol . 47 :247 .
14 . COTRAN, R. S ., and M . LITT . 1970 . Ultrastructural localization of horseradish peroxidase and
endogenous peroxidase activity in guinea pig
peritoneal macrophages. J. Immunol. 105 :1536 .
15 . ROBBINS, D ., R. S . COTRAN, and H . D . FAHIMI .
1971 . Fine structural cytochemical localization
of peroxidase activity in rat peritoneal cells .
Mononuclear cells, eosinophils and mast cells .
J. Histochem . Cytochem. 19 :571 .
WIDMANN ET AL .
16. FAHIMI, H . D . 1967. Perfusion and immersion
fixation of rat liver with glutaraldehyde. Lab .
Invest . 16 :736 .
17. FAHIMI, H . D ., and P . DROCHMANS. 1965 . Essais
de standardisation de la fixation au glutaraldéhyde . 1 . Purification et détermination de la
concentration du glutaraldéhyde . J. Microsc.
(Paris) . 4 :725 .
18 . SMITH, R . E ., and M . FARQUHAR . 1965 . Preparation of nonfrozen sections for electron
microscope cytochemistry . Sci. Instrum . News .
10:13 .
19 . GRAHAM, R. C ., and M . J . KARNOVSKY . 1966 .
The .early stages of absorption of injected
horseradish peroxidase in the proximal tubule
of the mouse kidney. Ultrastructural cytochemistry by a new technique . J. Histochem .
Cytochëm . 14 :291 .
20. CLAUDE, A. 1960. Problems of fixation for electron microscopy. Results of fixation with
osmium tetroxide in acid and alkaline media .
Pathol. Biol . 9 :933 .
21 . LUFT, J . H. 1961, Improvements in epoxy resin
embedding methods. J. Biophys. Biochem .
Cytol. 9 :409 .
22 . COTRAN, R . S ., E . R. SUTER, and G . MAJNO.
1967 . The use of colloidal carbon as a tracer
for vascular injury . Vasc . Dis . 4 :107.
23 . MAJNO, G . 1965 . Ultrastructure of the Vascular
Membrane. In Handbook of Physiology .
Williams and Wilkins Co., Baltimore, Md .
3 :2293 .
24 . ITO, T., and M . NEMOTO . 1952 . Uber die Kupfferschen Sternzellen and die "Fettspeicherungzellen" ("fat storing cells") in der Blutkapillarenwand der menschlichen Leber . Okajimas
Folia Anat . Jap . 24 :243 .
25 . BRONFENMAJER, S ., F . SCHAFFNER, and H . Popper. 1966 . Fat-storing cells (lipocytes) in
human liver . Arch . Pathol. 82 :447 .
26 . ITO, T., and S . SHIBASAKI . 1968 . Electron microscopic study on the hepatic sinusoidal wall
and the fat-storing cells in normal human liver .
Arch . Histol . Jap . 29 :137 .
27 . TÖRÖ, I ., P. RUSZA, and P . ROHLICH . 1962 .
Ultrastructure of early phagocytic stages in
sinus endothelial and Kupffer cells of the liver .
Exp. Cell Res. 26 :601 .
28 . MATTER, A ., L. ORCI, W . G . FORSSMANN, and
C . ROUILLER . 1968 . The stereological analysis
of the fine structure of the "Micropinocytosis
Vermiformis" in Kupffer cells of the rat . J.
Ultrastruct . Res . 23 :272 .
29 . AscHoFF, L . 1924 . Das Reticuloendotheliale
System. Ergeb . Inn . Med. Kinderheilk . 26 :1 .
30. LEwIS, E . H . 1931 . Pinocytosis.
Hopkins Hosp . 49 :17 .
Bull . Johns
Endogenous Peroxidase : Marker for Kupffer Cello
1 69
31 . FAWCETT, D . W. 1965 . Surface specializations of
absorbing cells . J. Histochem. Cytochem . 13 :75 .
32 . North,
R. J .
1970 . Endocytosis . Seminars
Haematol . 7 :161 .
33 . RABINOVITCH, M . 1970 . Phagocytic recognition.
In Mononuclear Phagocytes . R . van Furth,
editor, F . A. Davis Co ., Philadelphia, Pa.
313 .
34 . BRUNS, R . R ., and G . E . PALADE . 1968 . Studies
on blood capillaries . II . Transport of ferritin
molecules across the wall of muscle capillaries .
J. Cell Biol. 37 :277.
35 . KARNOVSKY, M. J . 1967 . The ultrastructural
basis of capillary permeability studied with
peroxidase as a tracer . J. Cell Biol. 35 :213 .
36 . VAN FURTH, R ., editor. 1970 . Mononuclear
phagocytes . F. A . Davis Co ., Philadelphia, Pa.
37 . PFUHL,
W.
1926 .
Experimentelle
Untersuchungen über die Kupfferschen Sternzellen
der Leber. Z. Anat. Entwicklungsgesch. 81 :90.
38 . PARKS, H . F ., and A . D. CHIQUOINE . 1956 .
Observations on early stages of phagocytosis of
colloidal particles by hepatic phagocytes of
the mouse . Electron Microsc . Proc . Stockholm
Conf. 154 .
39 . BROKELMANN, J ., and D. W. FAWCETT . 1969 .
The localization of endogenous peroxidase in
the rat uterus and its induction by estradiol .
Biol. Reprod. 1 :59.
40 . VENKATACHALAM, M. A ., M . H . SOLTANI, and
H . D . FAHIMI . 1970 . Fine structural localization of peroxidase activity in the epithelium of
large intestine of rat . J. Cell Biol . 46:168 .
41 . STRUM, J . M ., and M . J . KARNOVSKY . 1970 .
Ultrastructural localization of peroxidase in
submaxillary acinar cells . J. Ultrastruct . Res .
31 :323 .
peroxidase in rat exorbital lacrimal gland .
J. Histochem . Cytochem . 19 :216 .
43 . KLEBANOFF, S . J . 1967 . Iodination of bacteria : a
bactericidal mechanism . J. Exp. Med. 126 :
1063 .
44. KLEBANOFF, S . J . 1970 . Myeloperoxidase-mediated antimicrobial systems and their role in
leukocyte function . In Biochemistry of the
Phagocytic Process . Julius Schultz, editor.
American Elsevier Publishing Co ., New York .
89 .
45 . LEHRER, R . I . 1969 . Antifungal effects of peroxidase systems . J. Bacteriol. 99 :361 .
46. BELDING, M . E., S . J . KLEBANOFF, and C . G .
RAY . 1970 . Peroxidase-mediated virucidal
systems . Science (Washington) . 167 :195.
47. ORCI, L ., A . MATTER, and C. ROUILLER . 1971 .
A comparative study of freeze-etch replicas and
thin sections of rat liver. J. Ultrastruct . Res.
35 :1 .
48. ACKERMAN, G . A ., and M . A . CLARK. 1971 .
Ultrastructural localization of peroxidase
activity in normal human bone marrow cells .
Z. Zellforsch . Mikrosk . Anat . 117 :463 .
49 . BAINTON, D . F ., and M . F. FARQUHAR . 1970 .
Segregation and packaging of granule enzymes in eosinophilic leukocytes . J. Cell
Biol. 45 :54 .
50 . MILLER, F ., and V . HERZOG . 1969. Die Lokalisation von Peroxydase und saurer Phosphatase in eosinophilen Leukocyten während
der Reifung. Elektronenmikroskopisch-cytochemische Untersuchungen am Knochenmark
von Ratte und Kaninchen . Z. Zellforsch .
Mikrosk . Anat. 97 :84.
51 . STRUM, J . M ., and M . J . KARNOVSKY . 1970 .
of endogenous
Cytochemical localization
peroxidase in thyroid follicular cells . J. Cell
42 . EISNER, E. 1971 . Localization of endogenous
170
THE JOURNAL OF CELL BIOLOGY . VOLUME 5Q, 1972
Biol. 44 :655 .